Command of evidence: quantitative

What the exam asks

Each question stands alone: a passage of about 25–150 words (science, social science or humanities research) plus one graphic. You never need a calculator. At most you compare two numbers or notice which way a line moves.

Typical stem	What you must do
Which choice most effectively uses data from the table to complete the statement?	Fill the blank with data that makes the sentence true and on point
Which choice most effectively uses data from the graph to complete the example?	Supply a specific instance of the general claim just made
Which choice most effectively uses data from the table to support the researcher’s claim?	Find data that establishes the claim in the text
Which choice best describes data from the graph that support (or weaken) the hypothesis?	Find data that matches (or contradicts) a prediction

Easy versions ask you to locate one or two values. Hard versions, late in the harder module 2, make all four choices accurate and reward only the one that matches every part of the claim.

Core ideas

The two-test rule

Test	Ask yourself	Typical failure
Accurate	Does every number, label and direction match the graphic?	Wrong row, wrong year, reversed comparison, a value borrowed from the neighboring column
Relevant	Does it do the exact job the text needs?	True but about the wrong group, only half the claim, a snapshot instead of a trend

Check accuracy quickly by pointing to the cell or bar. Save your real thinking for relevance.

Find the claim before you read the graphic

The claim is almost always in the sentence with the blank or the sentence just before it. Break it into parts:

Who or what: which group, site, species or condition?
Which variable: temperature, pass rate, number of visits?
What relationship: greater than, increased over time, stayed stable, differed more in one condition than another, broke a pattern?
When: which years, trials or periods?

Only then look at the graphic, and only at the parts the claim names.

What each type of claim requires

Claim in the text	The correct data must show
Comparison (“X had more than Y”)	Values for both X and Y, in the right direction
Trend (“declined steadily”)	The start and the end (or every step), never a single value
Example of a general rule	One case from each side of the rule (e.g., a purple flower and a flower of another color)
Two-part claim (“an early advantage that later disappeared”)	Both parts
Interaction (“X helps more in A than in B”)	The effect of X in A and in B, with a bigger effect in A
Weaken	A case that breaks the predicted pattern

Reading graphics without errors

Title and units first: percent or raw count? Seconds or minutes?
Legend: in a grouped bar chart or a multi-line graph, decide which series is which before you read any value.
Line graphs: a rising line is an increase. Lines that cross show a reversal. Lines that converge show a shrinking gap.
Tables: say the row and column headers to yourself (“trial 10, crows that watched”) before quoting a number.

Worked examples

Completing an example (easy)

Urban ecologist Rosa Lindahl measured tree canopy cover and average summer surface temperature in four city neighborhoods. Her data suggest that neighborhoods with more tree canopy tend to be cooler. For example, ______

Neighborhood	Tree canopy cover	Average surface temperature
Elmwood	42%	31°C
Riverside	30%	33°C
Northgate	17%	36°C
Brickline	8%	38°C

Which choice most effectively uses data from the table to complete the example?

A) Northgate had 17% canopy cover, while Brickline had 8%.
B) Elmwood, with 42% canopy cover, averaged 31°C, while Brickline, with 8% canopy cover, averaged 38°C.
C) Riverside averaged 33°C, which was cooler than Northgate’s 36°C.
D) Brickline, with 8% canopy cover, averaged 31°C.

Solution. The claim links two variables, canopy and temperature. A gives only canopy and C only temperature, so neither shows the link. D is inaccurate, because Brickline averaged 38°C. B is accurate and pairs high canopy with a low temperature and low canopy with a high one. Answer: B.

Supporting an interaction hypothesis (medium)

Agronomist Devi Nair hypothesized that planting clover between rows of corn raises corn yields more on sandy soil than on clay soil.

Soil	Yield without clover (tons per hectare)	Yield with clover (tons per hectare)
Sandy	6.1	8.0
Clay	9.2	9.5

Which choice best describes data from the table that support Nair’s hypothesis?

A) On both soils, fields with clover yielded more corn than fields without clover.
B) Clay fields yielded more corn than sandy fields, with or without clover.
C) Clover raised yields on sandy soil from 6.1 to 8.0 tons per hectare but on clay soil only from 9.2 to 9.5.
D) Sandy fields with clover yielded more corn than clay fields without clover.

Solution. The hypothesis compares the size of clover’s benefit on two soils, so you need its effect on both. A is accurate but says only that clover helped on both soils, not that it helped more on sand. B is accurate but compares soils, not clover’s effect. D is inaccurate, because 8.0 is less than 9.2. Answer: C.

All four choices accurate (hard)

Historian Jonah Pell argues that the arrival of the railroad in 1870 transformed the town of Millbrook: factory employment grew rapidly after that year and soon made manufacturing the town’s largest source of jobs.

Year	Farm workers	Factory workers
1860	820	90
1870	800	150
1880	610	540
1890	480	910

Which choice most effectively uses data from the table to support Pell’s argument?

A) Farm employment was higher in 1860 than in any later year shown.
B) Factory employment rose in every decade shown.
C) Factory employment grew sharply after 1870, and by 1890 factory workers outnumbered farm workers.
D) In 1870, Millbrook had 800 farm workers and 150 factory workers.

Solution. Every choice matches the table, so decide on relevance. The argument has two parts: rapid growth after 1870, and manufacturing becoming the largest employer. B shows growth, including before the railroad, but never shows manufacturing overtaking farming. D is a single snapshot, and A is about farming alone. Only C covers both parts. Answer: C.

Exam technique

Read the stem first. Note whether you must complete, support or weaken.
Read the text and pin down the claim. Label its parts: who, which variable, what relationship, when.
Predict the kind of data you need (for example, “a purple flower beating a flower of another color”).
Skim the graphic for exactly those parts: title, units, legend and the relevant rows or lines.
Test each choice: Is it accurate? Is it relevant to every part of the claim? Eliminate at the first failure.
Pace yourself at 60–75 seconds. If two choices survive, reread the claim word by word. The difference is usually one word such as more, both, after or steadily.

Common mistakes

Quick recap

Every correct answer is accurate to the graphic and relevant to the claim.
Find and dissect the claim before you study the graphic.
Match the claim type: comparison, trend, example, two-part, interaction or weaken.
Read titles, units and legends first. Most inaccurate choices quote the wrong row or series.
On hard items all four choices may be accurate, so relevance decides.
No calculations: you only compare values and read directions.

By the end you’ll be able to

What the exam asks

Core ideas

The two-test rule

Find the claim before you read the graphic

What each type of claim requires

Reading graphics without errors

Worked examples

Exam technique

Common mistakes

Quick recap

Check your understanding

Flashcards

Command of evidence: quantitative

By the end you’ll be able to

What the exam asks

Core ideas

The two-test rule

Find the claim before you read the graphic

What each type of claim requires

Reading graphics without errors

Worked examples

Exam technique

Common mistakes

Quick recap

vii.Check your understanding

viii.Flashcards

Check your understanding

Flashcards